Comparison
Goodhart's lawvsVelocity
Goodhart's law
the team was measured on tickets closed, and the tickets got smaller.
When a measure becomes a target, it ceases to be a good measure. It is the central risk of every engineering metric — coverage, velocity, story points, incident counts, review turnaround — because the measure is a proxy and the proxy is easier to move than the thing. It is the reason the DORA metrics are published as a diagnostic set and the reason a good metric conversation includes how it would be gamed.
Full entry →Velocity
the number went up every quarter and the amount of software delivered did not change.
How much work a team completes per iteration, used to forecast how much the next one might hold. It is a capacity-planning input and nothing else; the moment it becomes a target it is trivially inflatable by estimating higher, which is Goodhart's law arriving on schedule. Comparing velocity between teams is meaningless because the unit is defined locally and deliberately.
Full entry →